English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

This paper studies the merits of applying log-gradient input images toconvolutional neural networks (CNNs) for tinyML computer vision (CV). We showthat log gradients enable: (i) aggressive 1.5-bit quantization of first-layerinputs, (ii) potential CNN resource reductions, and (iii) inherent robustnessto illumination changes (1.7% accuracy loss across 1/32...8 brightnessvariation vs. up to 10% for JPEG). We establish these results using the PASCALRAW image data set and through a combination of experiments using neuralarchitecture search and a fixed three-layer network. The latter reveal thattraining on log-gradient images leads to higher filter similarity, making theCNN more prunable. The combined benefits of aggressive first-layerquantization, CNN resource reductions, and operation without tight exposurecontrol and image signal processing (ISP) are helpful for pushing tinyML CVtoward its ultimate efficiency limits.

Improving the Energy Efficiency and Robustness of tinyML Computer Vision using Log-Gradient Input Images